Crystal is subject to numerous defects, sometimes present- ing a rough, solder-like, substance, or else clouded by spots upon it; whiie occasionally it contains some hidden humour within, or is traversed by hard and brittle inclusions . . . Some Crystal, too, has a red rust upon it, while, in other instances, it contains filaments that look like flaws. . . . Pli,ny, XXXVI, chap. 11.

THE II{SPECTION AND GRADING OF QUARTZ

Seuunl G. GonooN, AssociateCurator, Department of Mineralogy, The Academy of Natural SciencesoJ Philad.elphia.

CoNronrs Page Abstract.... . 269 rntroduction . ::: ..:. :. : : ::... ::.:..... : . :::: . 26s TwinninginQuartz ...... 271 Dauphin6 ("Electrical") Twinning . . 271 Brazil ("Optical") Twinning . . . 275 Twinning According to the Combined Laws. 278 Detection of "Optical" Twinning in Polarized Light .. 279 The Quartz "Inspectoscope" . .. . . 28r Field Model. 282 Detecting Optical Twinning in Practice. 284 Flaws and Inclusions in Quartz. 284 Grading Quartz. 287 Inspection During Processirg. . ... 288

Ansrnect

To be usable for radio frequency oscillators, a quartz crystal must have part of its volume free of twinning and detrimental inclusions. Dauphin6 ("electrical") twinning may sometimes be detected upon the surface of the crystal, or by etching crystals or sections. Brazil ("optical") twinning, intergrowths of right and left quartz, is easily detected by viewing a crystal, immersed in oil, and between crossed polaroids, in the direction of the opticaxis.An oil bath, so set up, has been calledaquartz inspectoscope. It also permits the quartz to be scanned in an intense beam of light for cracks, bubbles. inclusions, and the inexplicable oriented linear clouds in V-shaped clusters called "blue needles." The apparatus is not only useful in grading quartz for "usability," but for inspection of sections, bars, wafers, and blanks to avoid uselessprocessing. A portable field model, using sunlight, is described. An example of the rare type of twinning according to the "combined laws" is illustrated. fNtnooucrrow The usefulnessof quartz for radio-frequency oscillators depends upon the absenceof deleterious matter and structures that afiect its elasticity or causeelectrical anomalies.

269 SAMUEL G. GORDON

Frcs. 1-3. Typical Braziliancrystalsof radio grade. Fig. 1:a"candle"l Fig.2: acrystal with a large s (1121) piane, showing characteristic striationsparallel to edge s:r; Fig.3: a Dauphin6 ("electrical") twin, showing the characteristic sutured boundary on zz (10T0). INSPECTION AND GRADINGOF QUARTZ 271

We are not particularly concerned with the obvious inclusions of rutile, tourmaline, hematite, pyrite, mica, chlorite, or moving bubbles that are of interest to mineralogists but with quite common defects that occur in eye-clearrock crystal. Twinning is as prevalent in quartz as it is in plagioclase feldspar, and quartz suitable for cutting flawless crystal balls may consist of polysynthetic twinned laminae that destroy its use- fulness for radio purposes.Limpid quartz is apt to be filled with invisible bubbles, or the remarkable linear clouds in V-shaped clusters called "blue needles.tt Quartz oscillators may be cut from either left or right quartz, but the presenceof both in the same plate results in electrical anomalies. Since qtartz oscillators are cut in a definite crystallographic orientation, a plate oriented with respect to one part of a twin will be misoriented for the other. The presence of some minute bubbles and "blue needles" may be tolerated in large plates vibrating at low frequency (such as X-cuts and filter plates), but it is obvious that a bubble in a small plate will occupy an appreciable volume of the whole plate and may seriously aftect its elastic properties. It is now the general practice to etch oscillator plates to their nominal frequency-thicknessl bubbles and "blue needles" are rapidly attacked by the etching solutions, and their absenceshould be assured. It is not only necessaryto gradethe quartz on purchase,but to inspect it at various stagesof manufacture to avoid uselessprocessing. It is also possible to plan the cutting operations to get a maximum yield from twinned q\artz, and rules to accomplish this will be given in an accom- panying paper.

TwrNwrwc rN Quanrz Untwinned qnartz is quite rare, and it is unusual to find a crystal which is not twinned according to both the Dauphin6. and Brazil laws. Dauphinl ("electri.cal")Ltui.nning is the designation of Dauphin6 twins by radio engineers: an intergrowth of either two right or two left in- dividuals, at 180" to each other, with c as the twin axis (Figs.4-8). The rhombohedralfaces r (1011) and z (0111) become coplanar, as do also m (1010) and rn (1010). Such twins are often recognized by the sutured boundary on m (1010), or by a difference in the sheen of the coplanar forms r and z on the apex faces (Fig.9). Dauphin6 twins are also recognizableby the presenceof s (1121),and o (5161)on both edges of a prism face, when the normal, untwinned trigonal symmetry calls for these faces only on alternating prism edges.This type of twinning is 272 SAMUEL G. GORDON

Frcs. 4-8. Parallei growth and Dauphin6 (,,electrical',) twinning. Fig. 4: a parallel growth; Fig. 5: a Dauphin6 twin: an intergrowth of two right or tworeft crystals with the polarity of their X-axes reversed; Fig. 6: an idealized intergrowth of the individuals of a Dauphind twin showing r and z as coplanarl Fig.7: a more natural representation; Fig. 8: ap- pearance of an etched wafer showing the characteristic "ripple" and "shingle" structure developed by coplanar z and r, respectively. INSPECTION AND GRADING OF QUARTZ 273

Frc. 9 In Dauphin6 ("electrical") twinning, r (10T1) and: (0111) are coplanar, but oftendistinguishableby difierences insheen,due to naturaletching. (Photographthrough the courtesy of Lieut. William B. Gray, Jr.) called electrical twinning becauseit results in a reversal of the polarity of the electricalaxis (X [1120])at the twin boundary' In marked contrast to twins of the Brazil law ("optical twinning") the twin boundaries of Dauphine twins are quite irregular, and this servesto distinguish the two in etched sections(Figs. 12-13).1In Dau- phin6, ,,electrical" twins, the coplanar areas,r and z canbe distinguished in etched wafers by their difference in translucency and structurel r developes a characteristic "shingle" Structure, while z develops a less symmetrical ,,ripple" pattern (Fig. 8); this is further discussedin the following paper.

l Leydolt, Franz, Uber eine neue Methode, die Struktur und Zusammensetzung der Krystalle zu untersuchen, mit besonderer Beriicksichtigung der Variet?iten des rhomboed- XV (1) 59-81' rischen Quarzes.Sitzber. d. hais. Ahad,.d.Wiss.Wien (Math.-Natur. Classe) 1855. ar^ SAtrIUEL G. GORDON

Frcs. 10-11. Two views of a Dauphin6 (,,electrical',) twinned crystal, showing a sutured boundary on the prism m (lor0), and the equal development of two apex faces and zonal development of these with z (10T0). The latter, resulting in a pseudo-prismatic appear- ance of the apex faces with long edges in contact with zz is diagnostic in detecting electrical twinning. (Photographs by NAP.) F};

Frcs. 12-13. Etched sections reveal twinning strikingly, and Dauphin6 (,,electrical,') twinning can be distinguished from tsrazil ("optical") twinning by the irregular boundaries developed in the former (Fig. 12) in contrast with the straight boundaries developed in optical twinning (Fig. 13). These sections are perpendicular to the optic axis. (After Leydolt.t) INSPECTION AND GRADING OF QUARTZ 275

A Dauphin6 twin may consist of a ;iair with an irregular boundary between them. Other crystals contain many quite irregular patches- which suggests that the twinning may be secoudary; that is, due to stressesset up through sudden changes of temperature and pressure. A small plate heated in an oven (not necessarilyto the inversion point) will twin on sudden cooling, and etching will reveal curved boundaries sym- metrical to the edgesof the plate. Brazil "optical twinni.ng" is the name applied to twinning after the Brazil law, because it is readily detected by examination in polarized light. This kind of twinning is an intergrowth of the two enantiomor- phous forms: of right and Ieft individuals.2 The story of the discovery of this kind of twinning is recorded by Sos- man:3

"Twinning in quartz was first placed onrecord by Weissa in 1816.He described and figured clearly the orientational (Dauphin6) type of quartz twins. The chirai (Brazil) type he appears also to have observed, though rarely, but spoke of it cautiously as a kind of intergrowth which might possibly be explained as a case of twinning allied totheorienta- tional. The difiiculty was simply that the difference between the two alternative choices, -positive versus negative forms on the one hand, and right versus left on the other,-had not been made clear. "Herschels happened upon some twinned crystals in the course of his pioneer work on the correlation of optical activity and face development in quartz and was puzzled to find 'tending one specimen which had trapezohedral faces perfectly distinct and in contact but opposite ways around the summit.' Unfortunately it was'in the possession of Mr. Brooke,' and Mr. Brooke apparently valued it too highly to let it be damaged in the interest of science, so Herschel was not allowed to section it and test its rotatory power for polarized light. The sympathy of every true mineral collector wili be with Mr. Brooke, even though it did postpone an important scientffic discovery for twenty-five years. "The explanation for this peculiar crystal was not forthcoming until G. Rose6 pub- lished his comprehensive work in 1846. He was the first to show that the apparently homo- geneous crystal with the opposite tending faces is really a compound crystal, twinned ac- cording to the Brazil law."

2 Lord Kelvin (Baltimore lectures, 1904, p. 640) called it chi.ral twinning (from the Greek, for hand). The optical rotary power due to handedness has been called.chi.ral'ity. 3 Sosman, Robert B., Tke Properties oJ Silica, p. 196-197, Chemical Catalog Co., New York (1927). a Weiss, C. S., Uber den eigenthiimlichen Gang des Krystallisations-systemes beim Quarz, und iiber eine an ihm neu beobachtete Zwillingskrystallisation:Ges. Nat. Freunilc Mag. Berli,n 7, 163-181, pl. 4 (1816). 6llerschel, J. F. W., On the rotation impressed by plates of rock crystal on the'plane of polarization of the rays of light, as connected with certain peculiarities in its crystal- lization: Trans. Cambridge Phil. Soc. l,43-52 (1821). 6 Rose, Gustav, Uber das Krystallisationssystem des Quarzes: Physik. Abh. Kiingl'. Akad. Wiss. Berlin, T8M:217-274, pl. 1-6 (1846). 276 SAMUEL G. GORDON

Frc. 14. Brazil ("optical") twinning is the intergrowth of right and left quartz. It consists of laminae of one hand enclosed within a host crystal of the opposite hand. The remarkable shapes assumed by these "satellites" are sholvn in the ideaiized perspective of the Abbe C. Gaudefroy,T who listed as the principa\ forms m(ez) 1010, r(2) 1011, z(e1t2) 0111,s 1121,and c(ol) 0001. i ? Gaudefroy,C., Sur les grouppments de cristaux de quartz i axes paraliBlo.: Bull. soc. Jrang.min.56,5-63 (1933). { Lamellar twinning was discussed by John W. Judd: Additional note on the lamellar structure of quartz crystals, and the methods by which it is developcd: Min. Mag. lO, r23-r35(r89D. Similar twinning in cinnabar was described by Waldemar Lindgren: U. S. Geol. Surtt. Bull..6r (1890). INSPECTION AND GRADING OF OUARTZ 277

Crystals twinned according to the Brazil law and resembling the ideal- ized drawing of Rose, which has been copied in all text-books, are rare, and their existencerequires confirmation. This type of twinning is seen rather as a growth of laminae of one hand within a host crystal of the

Frcs. 15-19. Brazil ("optical") twinning may be detected by viewing crystals (im- mersed in oil between crossed polaroids) in the direction of the optic axis. The contour of the base or termination is seen in bancls of color expressing various degrees of rotation of the planeof pnlarizationbecauseof varyingthickness olqtartz(Fig. 15).Laminaeof theop- posite hand are seen as triangular areas (Figs. 16-17). Etching of sections Iocalizes the twin- ning (Figs. 18-19). (Photographs 15, 18-19 by NAP.) opposite hand (Figs. 14-19). The shapesassumed by these inclusions are remarkable: they suggestcrystals of the pedial class of the triclinic sys- tem, in that they are triangular polygons, and often with but one form 278 SAMUEL G. GORDON of a kind. Not only are they of a habit entirely different from that as- sumed by the host crystal, but they may show a well developed basal plane! The form of these "satellites" was described by the Abb6 C. GaudefroyT in 1933, then president of the Soci6te Frangaise de Miner- alogie, and the principal types are reproduced in his idealized perspective (Fig. 1a). As a rule the laminae are quite thin, and many laminations of apparently the same dimensions may occur rhythmically spaced above one another in the host, and invariably parallel to the major rhombo- hedral faces(r 1011)(Fig. 17). Most quartz crystals are intimate intergrowths of polysynthetic lam- inae of right and left quartz, and it is principally in Brazil that quartz crystals occur relatively free of the twinning to which that country has given its name! Quartz of radio quality from Brazil usually shows the parasitic laminae at the margins of the crystals, and the interior may be quite free of them.

Twinn'ing According to the Comb'ined.Laws Electrical twins (Dauphin6 law) are described as two right, or two left individuals, at 180oto each other with respect to the theoretical twin plane 1120;the faces r (1011) and z (0111) are coplanar.Optical twins (Brazil law) are simply intergrowths of the two left and right enantio- morphous form: the forms r, z, ?n on one individual are parallelto r, z, and m of the other. An example of the combined lawss is shown in Fig. 20. Two parallel planesperpendicular to the electrical axis (a, X-axis) have been cut on the crystal; one was etched and the other polished: the photograph shows the view looking through the polished (f) X-plane to the etched (-) X-plane at the back. From the parallelogram light figures (described in the accompanying paper on Orientation), the two parts may be consideredas an electrical twin of a right and a left crystal, or as an optical twin in which the right and left crystals are 180oto each other, with r and z coplanar. ft is characteristic of electrical twins to have irregular boundaries, while optical twins have straight boundaries: it will be noted that this specimenshows both; the boundariesparallel to r (1011) are plane, as it should be in an optical twin: the other boundary is quite irregular.

8 Found by Mr. Isaac Herr of the Commercial Crystal Company of Lancaster, Pa. Other eramples, discovered by Mr. Hoyt Brubaker, were generously given to the writer by Mr. P. R. Hoffman of Carlisle, Pa., for study. L. A. Thomas, Nalure, 155,425 (Aprrl7, 1945) suggests the name "compound optical twinning" for the combined laws. INSPECTION AND GRADING OF OUARTZ 279

DBrocrror.I otr ,,O?TrcAtrt TwrNuNc rw Poranrzpo Lrcnr Twinning according to the Brazil law is called ,,optical,, twinning be- cause it is so readily detected by viewing the quaftz between crossed polaroids in the direction of the optic axis. This may be readily demon- strated by immersing a clear qtartz crystal in a beaker or culture dish filled with an oil of proper index. The beaker is set upon a pair of glass

Fro. 20. A crystal twinned on the rare "combined Laws." view perpendicular to elec- trical axis [1120]: through the polished face [fxJ to etched face in back [-x]. sketched in are the light figures seen on the etched surface on viewing a point source of light. The central area has the character of both an optical twin (straight boundaries parallel to z (10T1)), and of an electrical twin (the irregular boundary). About2/s natural size. plates between which is a film of polaroid (a difiusing ground glassplate is also desirable). The plates are placed upon a laboratory tripod-an electric light within the tripod may be used for illumination. A second polaroid, crossedto the first, is set upon the beaker. Upon viewing the quartz crystal at a slight angle to the optic axis, the twinned laminae will be seenas thin, spectacularlycoloredplates (Figs. 15-17).Thephe- 280 SAMUEL G. GORDON

Frcs. 21-24. The quartz inspectoscope is simpiy an oi1bath in which a quartz crystal can be viewed between crossed polaroids to detect optical ti,r'inning; it also permits scanning of the crystal in an intense beam of light for flaws and inclusions. Fig.2l: the Klieg model used in the plant: the tank (T) has a pair of crossed polaroids (P), a light source (I) and a view- INSPECTION AND GRADINGOF QUARTZ 28I nomena is striking only if viewed within a few degreesof the optic axis, and it is essential that the quartz be immersed in an oil approximating the indices of refraction oI quartz.The most satisfactory liquid is ortho- free tricresyl phosphate (z: 1.55) sold under the trade name of Lindol.s In general, the laminae have the appearance of right triangles: with the base of the triangle parallel to the prism face with which it is usually in contact; the hypotenuse parallel to a prism face at 120" to the first (parallel to an m: r edge, or an X axis) ; and the third side parallel to an r:r ed.ge(Y axis; prism normal). Other plates appear as equilateral tri- angles, with all three sides parallel to traces of prism faces (m:r edges)' The usefulnessof this method dependsupon the fact that the parasitic laminae are thin and wedge-like,and they rotate the plane of polarization in the opposite direction to that of the host. The polysynthetic twin laminae are thin enough to show several orders of interference colors, while the host is so thick that interference colors are seenonly as colored bands fringing the base of the crystal and expressingvariations in thick- nessof the contour of the latter (Fig. 15). The interference colors shown by various thicknessesof quartz are tabulated in Table 1.10 ttINSPBctoscoPEtt Trro Quanrz The quartz inspection apparatus used in crystal-cutting plants is essentially a tank in which the quartz can be immersed in oil and exam- ined in polarized light, as well as in the intense beam of a spot-light (Figs. 2l-24). Often it was merely a fish aquarium, the edgesof which had been

s Obtainable from the Celanese-Celluloid Corporation, 180 Madison Avenue, New York City, for about $0.285 per pound in 5 gallon lots. Specify ortho-Jree,a grade which seems to be free of dermatological hazards. After immersion, crystals are readily cleaned with laundry soap and water. 10More than a century ago, Fresnel showed that quartz was birefringent in the direction of the optic axis. sosman quotes for o and e, respectively 1.5M1887 and 1.5442605, or a birefringence of 0.0000718. Plane polarized light traveling in the direction of the optic axis is separated into two circularly polarized rays, of opposite sign and different speeds, which emerge to form plane polarized rays with the interference color composed of the resultants oi th" ruy. which were not extinguished. Airy showed that rays not parallel to the optic axis were transmitted as two sets of elliptically polsrized waves. G. Szivessy (Fort. Min. Krist. Petr.2l, tll-168 (1937)) has s-hown that d qutattz is optically inactive at 56"10' to the optic axis. Rays traveling perpendicular to the optic axis are plane polar- ized. Full discussions of the rotatory power of quartz will be found in the works of Sosman and Tutton.

(Photographs,courtesy of NAP.) 282 SAMUEL G. GORDON

sealed with asphalt paint, and which had been fitted with screens of polaroid at the ends, similarly sealedagainst seepageof oil. illumination was provided by a pair of 40 watt electric light bulbs, difiused by ground glass or opal glass at one end, and a mirror was provided {or viewing the quartz at the other end.

Taslr polanrzArron 1. Rorarroll ol,rur PlaNe ol or eulnrz*

Thickness Thickness Color Fraun Rotation fnterference ave lengt necessary necessary extin- hofer per mm. color Linpp to rotate to rotate between guished line at 20"C. crossed 900 180" nicols

Red A 760.4 12.65" /-1.) 14.30 Green a 718.36 14.30 6.29 12.58 Blue B 686.71 15.75 J./I tr.42 Orange c 656.2r t7.31 5.19 10.39 Yellow DI 589.513 2t.69 4.14 8.29 Violet Dr 588.912 2r.725 4.tr 8.23 Green E 526.9r3 1i

+ Johannsen, A., Manud oJ Petrographi.e Methods, p. 109. McGraw-Hill, New york (1e14).

Such equipment was modeledafter the quaftz,,inspectoscope"devel- oped by the Bell Laboratories for the Western Electric Company, where mercury vapor lamps were used for illumination, and the light was filtered through a seriesof corning color filters including an rlusion pink, a yellow shade yellow, and a Light Shade blue green, as well as a heat filter. This instrument had advantages over direct viewing tanks used in the Bureau of Standards, where the lower polaroid plate was in the vulnerable position of being at the bottom of the tank and was soon ob- scured by dirt or cracked by quartz crystals slipping from the operator's hand: moreover, the mirror image seemedto present a better picture of the twinning that could be gathered from scanning the crystal directly. Field. Mod,el A smaller, portable apparatus (Fig. 22)rrwas introduced for use in the field, but proved equally popular in crystal-cutting plants. It was de- signed for use in regions remote from electricity, and was illuminated by 11 Manufacturedby Kliegl Brothers,321 West 50th Street,New york City: a Ktieg s_potlightis providedfor plantuse. INSPECTION AND GRADINGOF QUARTZ 283 light reflected from the sky and the sun by a pair of folding mirrors: for that reasonthey becameknown in Brazil as "sun buckets." A tight fitting Iid permitted transportation of the oil in the bucket. Upon reaching a prospect, the tank is set upon a rock, with the sun to one side of the prospector (Fig. 22). The lid is unclasped and set up as a shield to exclude extraneouslight. The mirrors are pulled out and turned

Frcs. 25-26. Eye-clear, limpid quartz is quite apt to show phantoms or "blue needles" on scanning the crystal, immersed in oil, in an intense beam of light. to their reflecting angles. The light of the sky overhead or in front of the prospector is reflected into a polaroid window, and the mirror opposite it is adjusted for viewing quartz for optical twinning. Another, circular, mirror is set to reflect the sun's rays into the tank for examination for flaws, bubbles, and other defects: a metal shield hung over the tank can be slid over this window to exclude the direct beam durine the examina- tion in polarized light. 284 SAMUEL G. GORDON

DetectingOptical Tuinning in Practice The quartz must be thoroughly cleaned before inspection; moreover cloudy areasat the sidesor basemust be trimmed by meansof a hammer, rock trimmer or saw: this is particularly true of such crystals as occur at Hot,Springs,Arkansas in which the lower part is apt to be quite cloudy. The clear crystal is held in the oil bath, with the optic (Z) axis direc- tion (c), perpendicular to the crossed polaroids. The irregular base will be fringed by concentric color bands expressingvariations in thickness of the margin of the crystal (Figs. 15-17). Twin laminae will appear as colored triangwlar blades along the side or shooting through the crystal, or cutting across the concentric color bands of the margin of the mass. By slightly tilting the crystal the position of the twinned areascan be ap- proximately located and usable volume of the quartz, ftee of twinning, can be estimated. Since it is necessaryto orient the qtartz precisely in order to see the twinning, somedifficulty may be experiencedat first with defacedquartz; the following rule will be helpful: Turn the mass horizontally in the oil bath until the color bands concentric with the margin appear or the qvartz is at extinction; in the latter case,rotate it now about an axis parallel to one of the polaroid vibration directions.

Flews aNn IucrusroNs rN Quanrz An intense beam of light from a carbon-arcor a projector is used to detect flaws and inclusions in quartz. The quartz is scanned while it is submerged in an oil approximating it in index of refraction, and the effect is as revealing as the beam from a moving-picture projector through a dust-laden darkened room. The principal defects are fractures, cavities, bubbles, "ghosts," and the remarkable "blue needles"and "blue feathers" (Figs. 26-28). Vari- ous designationsused by quartz inspectorsare: (1) "Bubbles": bubble-like cavities, distinguished as "large" or ttfine.i' ttVeils,tt (2) "heavy" or "fine": more or lesscontinuous sheets of small bubble-likecavities. (3) " Clouds" and "haze' ' : aggregatesof fine, bubble-like cavities. (4) "Ghosts" are the familiar "phantoms" of mineral collectors (Fig. 25): earlier growths withiir the crystal, which become visible when a beam of light is reflected from the minute fractures or parting planes which outline them. They are also seenas alternating colorlessand smoky zoneswithin the crystal. INSPECTION AND GRADING OF OUARTZ 285

Frcs.27-28. "Blue needies" or "blue feathers" are quite common inclusionsof linear clouds in V-shaped clusters, with a preferred orientation of the limbs of the V parallel to anr:z intersection edge, and the plane of the V parallel Io anr iace. The apex of the V always points to the base of the crystal, never to the terminated end. The reticulation is only apparent, the V's being at successive, often equidistant, levels. They seem to be sys- tems of minute parting planes. (About 3f4natural size.) The relation of the basal view (FiS. 28) to Fig. 27 is as follows: the edge which is common to the two is in contact, and the exact relation can be seenby folding Fig. 28 under Fig. 27 along this edge. (The prism face and basal section shown were kindly polished by William Van Horn.) 286 SAMUEL G. GORDON

(5) "Blue needles"and "blue feathers" (Figs. 26-28): theseare inex- plicable linear clouds which appear bluish becauseof selectiveabsorption of -rays of others colors. They occur as pairs of needlesforming a V, with the apex of the V always towards the base of the crystal, and the two arms of the V have a preferred orientation parallel to rlz edgesl in other words, a V outlines a plane parallel to z towards the base of the crystal.l2 The "blue needles" are not reticular aggregates, such as shown by rutile inclusions in rose quartz, although the photographs give this illu- sion through projection on the same plane of V's at successive,and often equally spaced levels below the surface. "BIue needles" are not uncom- mon in quartz, but patience and alertness are necessaryin hunting for them, since they become visible only in certain orientations in a power- ful beam of light while immersed in oil: for example, the basal view shown in Fig. 28 showed the blue needles only when the beam of light struck a prism face above a major rhombohedral face r (1011). Under a magnifi- cation of 30x, using a binocular microscope, the V is seen to include a number of finer needles radiating from the same point but not in the same plane. These finer needlesare seento be very narrow planes with a cross fiber structure: a single needle may be the intersection of two such planes. The fibers seem to be parallel to parting planes. Jutting out from the needles,at almost regular intervals; and more or Iessat right angles to the needles, are flat spear-heads: these seem to have a central line, which may have been a fiber, and a structure which appears to be made -up of triangles or rhombs. They diffract light into the circular dots seen gn the needlesas "dew-drops" in the oil baths. These structures are hard to resolve,even under a magnificationof 150X, but they seemnothing more than systems of linear parting planes. This interpretation is sup- ported by examinations of etched wafers, where the etching solutions have penetrated the blue needles:only linear cavities bounded by planes parallel to parting directions(r 1011) are seen(Figs. 29-32). The Portugese word "chuva" meaning a fine rain is said to be applied to "white needles" with "small bubbles" giving the appearance of dew- drops along a thin fiber.l3 Tynd.all efect is a bluish smoky appearanceof a beam of light passing through qlrartz, and probably due to the light scattering effect of ex- tremely fine blue needles.

12This is obvious from the angle of 45' shown by the V's oI fig. 27 as projected on the plane of the photograph. lsWillard, G. W., Raw quartz, its impurities and inspection: Bel,l System Technical J our., 22, 338-361 (1943) ; M ineral. Abstr acts, 9, 83 (1944). INSPECTION AND GRADING OF QUARTZ 287

Frcs.29-32. Photornicrographs by Dr. Thomas S. Stewart (magnification 7X,7:X, 22X, and 70X respectively), of a wafer which had been etched. It is assumedthat the etch- ing fluid has proceeded along "blue needles" which occurred in the quartz, because of the v- shaped clusters shown (Fig.29). Fig.30 shows the limbs of the v proceedingfrom the edgds of the triangular-pyramidal etch pits on the opposite side of the wafer. Fig. 31 emphasizes how the solutions followed parting planes in the crystal. Fig. 32 shows the ribbonJike, linear character of the etched planes; under 150X magnification one can see optical twin boundaries on these planes!

GnaorNc Quanrz To justify cutting, a qtartz crystal must have some of its space free of twinning and detrimental imperfections. Its value will depend on how much of it is truly flawless. Efiorts have therefore been made to grade qtrartz on the basis of its usability: defined as the estimated partial vol- ume that might yield oscillator-plates. 288 SAMUEL G. GORDON

Grading has heretoforebeen limited to the results obtained in the oil bath: on the basisof the amount of optical twinning found, and of such imperfections as is revealed in the intense beam of a spotlight. No effort has been made to grade quartz on the basis of the amount of electrical twinning present, since this would require etching of the crystals-and dealers object to anything being done to their quartz-in fact, the oil bath has already been too revealing. Weight classif'cati,on' Quartz is sold by weight, and is sorted into weights of 100 to 200 grams, 200 to 300 grams, 300 to 500 grams, etc. In the Bureau of Standardscode, the numbers l, 2, 3, etc. are used to designatethe lower limit in a box of quattz. Faced.quartz refers to material having at least one identifiable prism or rhombohedral plane measuring approximately f, inch square or more. Boulders and defacedqttartz are called unfaced qtattz' Optical quartz inchtd€sflawless material, 500 gra.msor more in weight, free of strain, but not necessarily free of electrical twinning. There are two gradesof usability, 45-6070 and 60-L00/6. Oscillator quartz is gradedla according to the estimated percentage of usable volume, as well as for the presenceor absenceof structures or de- fects that might limit its use for somepurposes. Grade 1 includes flawless material. Grade 2 contains as defects only blue needles,color or Tyndall effect. Grade 3 showsin the eye-clearportions someparting, bubbles, and inclusions. Grades I and2 are graded further according to percentageof usability, with limits of 30-45/6, +5-6070, and 60-100/6 (or roughly into more than a quarter, a half, or three-quarters usable). Grade 3 was classified into two gradesof "usability" of0-45/6 and45-I00/s.

IxsrncrroN Dunrwc PnocBssruc There are rnany steps in the reduction of a quartz crystal into the tiny oscillator blanks: the quartz musL first be sliced into wafers, or if large, into sections, and then into bars suitable for wafering. More than 95/e of the quartz is uselessor wasted in the processing,and the final oscil- lators may represent less than 1/6 oI the volume of the quartz that one started with. Grading of raw quartz is necessary to be sure that the quartz has flawless areas that justify its processing' When the quartz has been cut into sections,bars, and wafers it is the practice to etch

la National Bureau of Standards, Optics Division (FJB:MLH, IV-w Ipp-941-c) "Spec- ifications for testing and grading crystalline quartz in conformity with the conditions irnposed by the procurement, purchase, distribution, and use of the material as found !I theBoard of Economic warfare, Metals Reserve company, war Production Board, and the U. S. Army Signal Corps." Washington, D. C.1943- INSPECTION AND GRADING OF QUARTZ them, and reject hopelesslytwinned areas.The most important stages of inspection, however, are those of the blanks prior to lapping and final finishing. If cut from bars they should be etched and examined for twin- ning; it is assumedthat only untwinned blanks were diced from etched

Frcs. 33-35. Amongthemostimportant stages of inspection in processingare those of the blanks prior to lapping and prior to final finishing. Prior to lapping, blanks are in- spected for internal imperfections (cracks, bubbles, inciusions) by immersing the blanl